[krisjan]

oh and i just want to thank this tangentsoft guy for his wonderful site and information sharing - it's really great to get all this info and step-by-step help for free! cheers!

 

[millwood]

Quote:

would higher impedance phones have less hiss at higher source volumes?

hiss is almost a universal sign of opamp oscillating and you shouldn't have it at all if the amp is working properly.

that aside, a higher impedance phone means less attenuation by R5, everything else being equal, that and means louder sound on the headphone.

but phone sensitivity typically goes up for higher impedance phones so that's hard to say.

 

[krisjan]

Quote:

Originally Posted by millwood /img/forum/go_quote.gif hiss is almost a universal sign of opamp oscillating and you shouldn't have it at all if the amp is working properly. that aside, a higher impedance phone means less attenuation by R5, everything else being equal, that and means louder sound on the headphone. but phone sensitivity typically goes up for higher impedance phones so that's hard to say.

well i've been looking at Basic Troubleshooting for Headphone Amplifiers and i'm quite sure my problem is listed as Quiet hiss in the headphones...as i said the bsg cmoy also has it...

 

[millwood]

Quote:

Originally Posted by krisjan /img/forum/go_quote.gif well i've been looking at Basic Troubleshooting for Headphone Amplifiers and i'm quite sure my problem is listed as Quiet hiss in the headphones...as i said the bsg cmoy also has it...

and I am sure you are right about it,

.

 

[Pars]

Quote:

Originally Posted by krisjan /img/forum/go_quote.gif well i've been looking at Basic Troubleshooting for Headphone Amplifiers and i'm quite sure my problem is listed as Quiet hiss in the headphones...as i said the bsg cmoy also has it...

Drop your gain to as low as 2 or so (check opamp datasheet for gain that it will handle as some don't like below 5 or so). I usually used 10k R4 and 2.2k R3 for a gain of 5.5 and it seemed pretty universal, but I didn't have any IEMs.

 

[millwood]

Cmoy got the overall gain right but it was done via a 50% attenuation, achieved through R1/R2.

and R1 is indeed needed. you don't want the opamp's base / gate to see the signal directly. R1 is there to help with parasitics.

but R1 = 100K is just too big. 1K is a typical number, sometimes even smaller than that.

so a better way to achieve the same gain is to use the following:

R1=1K;
R2=R3//R4 - assuming C2 is in place;
R3=10K;
R4=whatever needed to achieve your gain. 10K - 33K is a good range.
R5=11 - 120ohm, depending on your headphone.

+ two decoupling caps, and an opamp with high current capability (>100ma in my view), you have one hell of a headphone amp.

 

[millwood]

Quote:

ideally, you don't want the amp's input impedance (=R2//opamp's input impedance =~R2 since opamp's input impedance is much larger than R2) to overwhelm the pot - but even if that happens it is not the end of the world.

some people asked me why letting R2 overwhelming the pot is not a big deal.

resistive loading of a pot is actually one of the oldest tricks in electronics. it dates back to the days when logrithmic pots were expensive vs. linear pots. when you resistively load a linear pot, the combination produces a near perfect logrithmic pot.

see the attached chart for an example where a 10k pot is loaded with a 1k, 10k and 100k resistor, respectively. As you can see, 100k loading almost retains the pot's original linearity. 10k loading isn't that bad but when you load the pot with a 1k resistor, the signal passed through the combination turns very much like the signal going through a logrithmic pot.

so if you are going to have a resistive loading on a volume pot, buy a linear pot, not a log pot. very counter-intuitive.

the curve holds for other resistor values, as long as the relative ratios hold.

the problems with resistively loading a pot?

well, it creates one potential issue: the input impedance changes with the pot's position: when the wiper is at the top, the input resistance is R2. when the wiper is at the bottom, the input resistance is Rpot.

for a 1k resistor and 10k pot, it goes from 1k to 10k. most preamps will have no problem driving a 1k load so in reality it isn't that bit of a deal.

 

[Pars]

It kind of works. Tangent did a writeup of shunting a linear pot

Notes on Audio Attenuators

But you are right in that the R2 resistor at 100k really puts you in a bind in terms of what you can do with R3/R4 in order to have equal impedances at both opamp pins.

 

[krisjan]

ok well - here it is - my first ever diy amp. It was as i suspected, once i inserted the pot and audio inputs the noise was gone - maybe with all the interference on those clamp test leads and open wires i picked up some noise on the first test run...i swapped out the opamp for a 2107 - dc measured fine - one channel basically on 0mv and the other around 10mv which is still well below the recommended 20mv, and it sounds great! listening to it now as i type...only thing i'd still like to do is add wallpower - the price of these 9v's are ludicrous plus then the power is steady - so eventually i'll just get a 24v supply and then i'll be happy. Here's a pic - lo quality snap from the cellphone, but you'll get the idea (please don't laugh at my overspec input caps - voltage rating on those a bit overkill lol):

i think the next one i'll try is the starving student amp...man i just wish it was easier to source parts here in my country - such a pain if you have to order from ebay or sth and then wait for weeks...

 

[nsx_23]

Does anyone here detect major differences between running a CMOY of 1 9V VS 9V batteries?

The reason I'm asking is that with my previous 2 builds, I couldn't get 2 9V batteries to fit in the altoids tins at all, and am wondering if there are any major negative effects.

I completed another build last night which uses 2 9V, and am now comparing it to the single 9V builds. The dual 9V seems to have slightly better bass so far, but more testing is needed.

And lol at the oversized caps above! I used 470uf caps and thought they were big enough...

 

[krisjan]

Quote:

Originally Posted by nsx_23 /img/forum/go_quote.gif Does anyone here detect major differences between running a CMOY of 1 9V VS 9V batteries? The reason I'm asking is that with my previous 2 builds, I couldn't get 2 9V batteries to fit in the altoids tins at all, and am wondering if there are any major negative effects. I completed another build last night which uses 2 9V, and am now comparing it to the single 9V builds. The dual 9V seems to have slightly better bass so far, but more testing is needed. And lol at the oversized caps above! I used 470uf caps and thought they were big enough...

those big caps are the input caps for removing incoming dc - they are 0.47microf which is not unusual, but what makes them big is the voltage rating - they are 400v...i just followed the link in rs-components site to a "recommended" subsitute for the discontinued part listed on tangents site and didn't check if there was a lower voltage rating one available - anyway, luckily space was not an issue for me.

I was also wanting to compare the diff between one 9v and two - i know that in my bsg cmoy quality did degrade quite a bit when the battery neared the end of it's charge - but aside from this opamps such as the 2107 can basically not run on a 9v since +-4.5v is the bear minimum specified - in my old bsg cmoy i use a 12v wallpower supply when i'm using that chip...briefly, and rather subjectively, i also feel that the bass is tighter and more responsive on the two 9v's...

 

[nsx_23]

Hmm, I'm planning on adding a DC jack to mine sometime this week.